Free 60-second pre-flight audit. We run the same 30+ checks Amazon does against your PDF — line by line — and email you a /100 score with every issue flagged. No payment, no signup.
Drop your manuscript PDF and we run the same 30+ checks Amazon does, line by line, in under a minute. You get a clear /100 score, every failure point flagged, and the exact fix for each one.
Whether you book us, do it yourself, or hire someone else — running this first means you don't pay for a service you didn't need.
"Will my book pass KDP review?"
Need an NDA before uploading? Request a mutual NDA
The same 30+ checks Amazon's KDP system runs on every file you upload — but you get to see the answer before they do.
Scales by page count. 0.375" gutter for 24-150pp, up to 0.875" for 700-828pp.
Every image checked against KDP's 300 DPI minimum. Flags upscaled stock photos.
Every font in the PDF traced to its embed status. Missing fonts = guaranteed rejection.
Every entry in the table of contents tested against its target. Broken links are common.
Page count must be even on print. Blank-page conventions for new-chapter recto starts.
ISBN in metadata vs ISBN on the copyright page. Single most missed cause of rejection.
If content runs to the edge, bleed must be 0.125". If it doesn't, bleed must be off.
Calculated from final page count and paper weight. Wrong spine = cover rejected.
Full EPUB 3.0 validator pass. Manifest, OPF, NCX, accessibility metadata.
Min 24 / max 828 (B&W white), 72-600 (standard colour). Trim-size specific.
Cover, interior and metadata all declare the same trim size. Mismatch = manual review.
Including footnote anchors, line spacing, paragraph-style consistency, copyright-page completeness, content/quality flags.
If your book is already on Amazon, these two are live and free right now.
Drop your Amazon URL. Cover at thumbnail size, title block on search, blurb opener, review base — out of 100 with a "ready / test small / not ready" verdict.
Run the scorecard →Live retrieval test against ChatGPT, Claude, Perplexity, Gemini and Amazon Rufus (simulated). Tells you what each model knows about your book — with web search active and from training alone.
Run the AI score →